【摘要】 决策系统的约简是Rough集理论的核心内容之一,本文提出了一种决策表约简的算法。该算法通过构造一个差异表,不需要对决策表中的记录进行逐条考察,在从差异表中选出重要属性同时也抽取该属性的对对象分类起决定作用的属性值,最后得到原决策系统的简化形式。与传统的约简算法相比,具有更低的复杂性。更多还原

【Abstract】 The reduction of the decision system is an important part in Rough Set Theory, so the paper proposed an algorithm which can reduce the decision table. Through constructing a Discernibility Table, the algorithm can extract the most important attribute and the relevant attribute value but does not have to inspect the objects one by one and then gains the reduced decision table. Comparing with the traditional algorithms, the algorithm proposed by this paper has the lower complexity.更多还原